Top

Overview

This report shows the baseline and robustness results from the empirical exercise in the paper “Do local and foreign newspapers unveil the same economic policy uncertainty shocks?” by E. Andres-Escayola, C. Ghirelli, L. Molina, J.J. Perez, and E. Vidal. In particular, here we report the impulse response functions from the Bayesian VAR model. Please refer to the paper for specific details.

Prepared by E. Andres-Escayola

Baseline results

Press comparison

IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Local press

IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

Spanish press

IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

International press

IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 90%-10%, 84%-16%, 68%-32% credible sets for 15 periods.

Robustness results

One lag

IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Cholesky permutation

IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Monthly data

IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Additional financial variable

IRFs to EPU shocks

Residential portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

EMBI & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods (filled triangles represent significant estimates).

Idiosyncratic and common factor

IRFs to EPU shocks

Portfolio capital flows & GDP (base)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Portfolio capital flows & GDP (idio)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Portfolio capital flows & GDP (com)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Exchange rate & GDP (base)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Exchange rate & GDP (idio)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Exchange rate & GDP (com)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Equity & GDP (base)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Equity & GDP (idio)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Equity & GDP (com)

(Percentage points)

Source: Own estimations.

Note: The baseline (base) results include the common factor and the idiosyncratic (idio) results exclude the common factor. The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. The local press corresponds to the red chart, the Spanish to the black one, and the Anglo-saxon to the blue one. We report the median response with the corresponding 84%-16% credible sets for 15 periods.

Country-specific results

Benchmark comparison

BR IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.

Robustness idiosyncratic comparison

BR IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.

Robustness common

BR IRFs to EPU shocks

Portfolio capital flows & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.

Exchange rate & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.

Equity & GDP

(Percentage points)

Source: Own estimations.

Note: The first chart corresponds to the response of the financial variable and the second to the response of GDP. Both responses are with respect to a 1 STD shock of the EPU. We report the median response with the corresponding 84%-16% credible set for 15 periods.